Robotic-assisted segmentectomy is a pulmonary resection procedure that is emerging as an alternative to lobectomy for the treatment of early-stage lung cancer tumours less than 2 cm in maximal diameter. Segmentectomy offers better lung function after surgery by only removing a few segments of the lobe that contain the tumour, and sparing remaining healthy lung tissue. As tumours are being more frequently detected in their early-stages, segmentectomy has gained considerable attention for its potential as a primary treatment option for suspected nodules less than 3 cm in maximal diameter. However, there is a reluctance in adopting segmentectomy due to technical challenges while performing the operation, and the lack of high-quality prospective data compared to lobectomy, which is the current standard of care. From a technical standpoint, segmentectomy is difficult to perform because the pulmonary lines that separate segments, or intersegmental planes, are invisible. This poses a challenge for the operating surgeon in determining where to resect the lung tissue to obtain adequate margin distance from the tumour. Near-infrared mapping (NIF) with indocyanine green dye (ICG) is a recent advancement in robotic-assisted segmentectomy that provides a complete delineation of the intersegmental plane. Previous work at our center has also shown that this technique was associated with an increase in the oncological margin distance compared to the surgeons’ initially estimated resection line. Given that segmentectomy is associated with a learning curve, we evaluated whether this was observed due to our early experience in robotic-assisted segmentectomy, and hypothesized that the added benefit of ICG would diminish as more cases were performed. In Chapter 2, we used a temporal analysis to monitor surgeon experience over time, and found that the clinical utility of NIF mapping diminished after approximately 42 cases with ICG, and the surgeon began to identify the location of the intersegmental plane more accurately and consistently without ICG injection since. The second barrier in the adoption of segmentectomy is the lack of high quality-prospective data. Current evidence pertaining to the effectiveness of segmentectomy in terms of cancer-related outcomes is inconclusive and difficult to generalize to the current lung cancer population. In Chapter 3, we performed a secondary analysis of a prospectively collected database of participants who underwent robotic-assisted segmentectomy or lobectomy for tumours less than 3 cm. The oncological efficacy of segmentectomy can be evaluated by the measuring the number of lymph node stations sampled intraoperatively and rates of nodal upstaging, and comparing these outcomes to pulmonary lobectomy. These are important surrogate outcomes that can be readily evaluated, and have been shown to predict overall survival after lung resection. We observed that these outcomes, including overall survival, were similar between patients who underwent segmentectomy and lobectomy for tumours less than 3 cm. While these findings were consistent for patients that underwent segmentectomy for tumours between 2 and 3 cm, recurrence-free survival was found to be significantly lower after segmentectomy compared to lobectomy. In conclusion, the clinical utility of near-infrared mapping diminishes over time, which is indicative of an improved ability to perform robotic-assisted segmentectomy as more cases were attempted. Second, adequate lymph node evaluation can be expected after segmentectomy, reducing the likelihood of missing positive lymph nodes. Although patients who underwent segmentectomy for tumours greater than 2 cm may be at a greater risk of experiencing recurrence compared to lobectomy, this population did not experience any reductions in overall survival. / Thesis / Master of Health Sciences (MSc)

Maskininlärning (eng: Machine Learning) har på senare tid blivit ett populärt ämne. En fråga som många användare ställer sig är hur mycket data det behövs för att få ett så korrekt svar som möjligt. Detta arbete undersöker relationen mellan inlärningsdata, mängd såväl som struktur, och hur väl algoritmen presterar. Fyra olika typer av datamängder (Iris, Digits, Symmetriskt och Dubbelsymetriskt) studerades med hjälp av tre olika algoritmer (Support Vector Classifier, K-Nearest Neighbor och Decision Tree Classifier). Arbetet fastställer att alla tre algoritmers prestation förbättras vid större mängd inlärningsdata upp till en viss gräns, men att denna gräns är olika för varje algoritm. Datainstansernas struktur påverkar också algoritmernas prestation där dubbelsymmetri ger starkare prestation än enkelsymmetri. / In recent years Machine Learning has become a popular subject. A challange that many users face is choosing the correct amount of training data. This study researches the relationship between the amount and structure of training data and the accuracy of the algorithm. Four different datasets (Iris, Digits, Symmetry and Double symmetry) were used with three different algorithms (Support Vector Classifier, K-Nearest Neighbor and Decision Tree Classifier). This study concludes that all algorithms perform better with more training data up to a certain limit, which is different for each algorithm. The structure of the dataset also affects the performance, where double symmetry gives greater performance than simple symmetry.

Call number: LD2668 .T4 1978 B42 / Master of Science

Virtual work environments (VWEs) have been used in the private sector for more than a decade, but the United States Marine Corps (USMC), as a whole, has not yet taken advantage of associated benefits. The USMC construct parallels the bureaucratic organizational culture and uses an antiquated information technology (IT) infrastructure. During an effort to upgrade the Marine Corps Combat Development Command's infrastructure to a VWE, the change-agent noticed an immediate resistance towards the VWE and new work methodology. The problem identified for investigation was to discover why a bureaucratic organizational culture, matured through IT savvy and cognitively adept personnel, resists the VWE and new work methodology introduced by the evolution of IT. The explanatory, single case study documented the resistance towards the VWE and new work methodology and recommended a solution to the problem. Due to a noticeable resistance geared towards the adoption of the VWE, the case study and pre-trial preparation began in Fall, 2009 and the data-collection period occurred in the Spring of 2010. The preparation phase entailed developing extensive instruments that burrowed into the participants' technical expertise and willingness to accept change on an individual level. The instruments were validated by an expert panel from the following disciplines: knowledge management, information technology, and psychology, as part of the groundwork. Analysis of the data showed resistance towards the VWE, both collectively as a group and on the individual level. The final report articulates that the data proves the study successfully accomplished the goal. Resistance was found at the user level in the work environment. It stemmed from several key areas: lack of user input, lack of training, user ignorance, and absence of a vision statement. Leadership should review the implementation methodology and decide upon a new course of action. Training in the use of the VWE at entry and advanced levels should be offered to newcomers and be available on a continuous basis. Future change agents must examine the outcomes of similar, previous work in order to gain a better understanding of what makes an initiative fail or succeed.

Nos recentes leilões de energia realizados no setor elétrico brasileiro, a energia térmica foi uma das principais vencedoras. Este trabalho avalia a possibilidade de reversão desta tendência, mantendo a trajetória anterior de uma matriz limpa e renovável. A maior parte da eletricidade brasileira tem sido proveniente de hidrelétricas. O plano energético oficial, com horizonte de 2030, elaborado para o Governo pela Empresa de Pesquisa Energética (EPE) dá ênfase à geração térmica, à gás natural, carvão e nuclear, como alternativa de complementação à geração hídrica. Neste estudo, em contraponto à proposta oficial, a geração eólica é analisada como opção de complementação, ao invés da energia térmica. A curva de aprendizado da tecnologia eólica, no Brasil e no mundo, é investigada e seu resultado evidencia o potencial de competitividade quando comparada a outras fontes, como térmicas nucleares, a gás e a carvão. A substituição do parque de expansão térmica pela eólica é simulada mediante a análise comparativa dos custos de capital, combustível, operação e manutenção, considerando a curva de aprendizado potencial. Os resultados da simulação, em termos de custo a valor presente das alternativas, indicam que a geração eólica pode se tornar atrativa, tendo como atrativo adicional a redução da emissão de gases de efeito estufa. Dificuldades e barreiras para a penetração da geração da energia eólica são avaliadas. Também é analisada a possibilidade do atendimento da demanda de energia no Brasil no contexto do cenário da estabilização populacional e do consumo, na década de 2040, mediante oferta de energia renovável, substancialmente hidro-eólica. / In the recent energy auctions held in the Brazilian electric sector, thermal power plants were the major winners. This study evaluates the possibility of reversing such trend, maintaining the previous path of a clean and renewable energy mix. Most of Brazil\'s electricity has been generated by hydropower. However the official energy plan, with the horizon of 2030, prepared for the Government by the Energy Research Company (EPE) gives emphasis to the thermal generation to natural gas, coal and nuclear energy as an alternative to hydropower generation complementation. In this study, in contrast to the official proposal, wind generation is considered as an option to complement, rather than the thermal energy. The learning curve of wind technology in Brazil and worldwide, is investigated and its result shows the potential of competitiveness compared to other sources such as nuclear thermal, gas and coal. The replacement of thermal based expansion by the wind is simulated, by a comparative analysis of capital costs, fuel, operation and maintenance, considering the potential learning. The simulation results in terms of present value cost of the alternatives indicate that wind generation can become attractive, with the added benefit of reduction in emission of greenhouse gases. Difficulties and barriers to the penetration of wind power generation are evaluated. The possibility of meeting the demand for energy in Brazil in the context of the scenario population and energy demand stabilization, in the 2040s, through renewable and sustainable energy sources, substantially hydro and wind, is also assessed.

INTRODUÇÃO: Este estudo analisou o impacto da experiência em prostatectomia radical robótica na curva inicial da prostatectomia laparoscópica através dos resultados perioperatórios e funcionais e oncológicos precoces de 110 pacientes com câncer de próstata. PACIENTES E MÉTODOS: 110 pacientes, selecionados aleatoriamente, foram submetidos a PRL por dois cirurgiões com experiência prévia em cirurgia minimamente invasiva e PRR, no Instituto do Câncer do Estado de São Paulo. Os dados foram coletados prospectivamente. Foi utilizada abordagem transperitoneal simulando a técnica robótica. RESULTADOS: Houve uma redução significativa no tempo operatório que diminuiu até o caso 40, quando atingiu um platô. O volume de sangramento mediano foi 250ml (variação 50-1000ml). Nenhum paciente foi transfundido. As complicações foram distribuídas uniformemente ao longo da casuística. A taxa de margens positivas foi de 28,2% (pT2: 20%, pT3: 43,6%), e não apresentou tendência a redução. 61,3% das margens foram apicais. A taxa de continência foi de 88% e atingiu um platô próximo a 95% após 75 casos. CONCLUSÃO: Esse estudo mostra que existem múltiplas curvas no processo de aprendizado da PRL. A curva mais curta foi para tempo operatório. A curva para margens cirúrgicas não foi completada e necessita de intervenção para melhora dos resultados principalmente relacionados a margem apical. A transição entre as técnicas de prostatectomia pode ser considerada segura baseada nas baixas taxas de complicações, ausência de complicações graves ou transfusões sanguíneas / INTRODUCTION: This study analyzed the impact of the experience with Robotic-Assisted Laparoscopic Prostatectomy (RALP) on the initial experience with Laparoscopic Radical Prostatectomy (LRP) by examining perioperative results and early outcomes of 110 patients. LRPs were performed by two robotic fellowship trained surgeons with daily practice in RALP. PATIENTS and METHODS: 110 LRP was used to treat aleatory selected patients. The patients were divided into 4 groups for prospective analyses. A transperitoneal approach that simulates the RALP technique was used. RESULTS: The median operative time was 163 minutes (110 - 240), and this time significantly decreased through case 40, when the time plateaued (p = 0.0007). The median blood loss was 250 ml. No patients required blood transfusion. There were no life-threatening complications or deaths. Minor complications were uniformly distributed along the serie (P= 0,6401). The overall positive surgical margins (PSM) rate was 28.2% (20% in pT2 and 43.6% in pT3). PSM was in the prostate apex in 61.3% of cases. At the 12-month follow-up, 88% of men were continent (0-1 pad). CONCLUSION: The present study shows that there are multiple learning curves for LRP. The shallowest learning curve was seen for the operative time. Surgeons transitioning between the RALP and LRP techniques were considered competent based on the low perioperative complication rate, absence of major complications, and lack of blood transfusions. This study shows that a learning curve still exists and that there are factors that must be considered by surgeons transitioning between the two techniques

In terms of the British Army in the Great War, the study of whether or how the army learned has become the dominant historiographical theme in the past thirty years. Previous studies have often viewed learning and institutional change through the lens of the 'learning curve', a concept which emphasises that the high command of the British Army learned to win the war through a combination of trial and error in battle planning, and through careful consideration of their collective and individual experiences. This thesis demonstrates that in order to understand the complexities of institutional change in the Great War, we must look beyond ill-defined concepts such as the learning curve and adopt a more rigid framework. This thesis examines institutional change in the British First Army in the 1915 campaign on the western front. It applies concepts more commonly found in business studies, such as organisational culture, knowledge management and organisational memory, to understand how the First Army developed as an institution in 1915. It presents a five-stage model - termed the Organisational Development Model - which demonstrates how the high command of the First Army considered their experiences and changed their operational practices in response. This thesis finds that the 'war managers' decision-making was affected by a number of institutional and personal 'inputs' which shaped their approach to understanding warfare. This thesis examines the manner in which new knowledge was created and collated in the immediate post-battle period, before studying how the war managers considered new information, disseminated it across the force and institutionalised it in the organisation's formal practices, structures and routines. In a broad sense, this thesis does three things. First, by examining how the army learned it moves beyond standard narratives of learning in the British Army in the Great War and highlights the complex interplay between personal and institutional learning processes. Second, by focusing on institutional change in the 1915 campaign, it sheds new light on an understudied yet crucial part of the British war experience. Finally, in creating the Organisational Development Model, it provides a robust platform on which future research can be built.

Introdução: A curva de aprendizado em prostatectomia radical laparoscópica (PRL) pode variar de 10 a 150 procedimentos. Nesta fase, observa-se o maior número de complicações perioperatórias e conversões, além de resultados oncológicos e funcionais precários. Neste estudo, foram comparadas duas séries iniciais de PRL, realizadas pelos acessos transperitoneal (PRLT) e extraperitoneal (PRLE). Objetivos: Comparar os resultados obtidos com a realização da PRL pelos acessos transperitoneal e extraperitoneal durante a curva de aprendizado, avaliando-se os resultados perioperatórios, oncológicos e funcionais. Pacientes e métodos: Procedeuse a uma análise comparativa retrospectiva entre os dados das primeiras 40 PRLT realizadas no Hospital Getúlio Vargas de Pernambuco (grupo 1) e os dados das primeiras 40 PRLE realizadas no Hospital das Clínicas da Faculdade de Medicina da Universidade de São Paulo (grupo 2). Resultados: Comparando-se as variáveis trans-operatórias dos grupos 1 e 2, observamos diferenças estatísticas na preservação dos feixes vásculonervosos (77,5% vs 90%; p=0,008), tempo cirúrgico total (175,0 min vs 267,6 min;p<0,001) e da perda sanguínea estimada (177,5 ml vs 292,4 ml; p<0,001). Duas complicações (5%) foram observadas no grupo 1 (sangramento e lesão retal) e quatro (10%) ocorreram no grupo 2 (sangramento-2, lesão retal e de bexiga). Conversão para procedimento aberto foi necessária em um caso em cada grupo (2,5%). No período pósoperatório, as principais diferenças estatisticamente significantes entre os grupos 1 e 2 foram observadas nas comparações do tempo de internamento, do tempo de uso de catéter uretral, no uso de opióides, na recorrência bioquímica, na taxa de continência urinária e no tempo médio de seguimento. Nenhuma diferença estatística foi observada na incidência de complicações precoces (17,5% vs 17,5%; p= 1,000), porém três complicações maiores foram observadas no grupo 1, levando ao óbito de um paciente neste grupo. Dentre as complicações pós-operatórias tardias, observou-se uma diferença estatística quando se comparou a taxa de complicações menores entre os grupos 1 e 2 (30% vs 15%; p=0,004). Na comparação dos resultados oncológicos entre os grupos 1 e 2, observou-se diferença estatística no número total de margens cirúrgicas positivas (MCP) (10,3% vs 32,5%; p=0,016) e no estadiamento patológico (pT2: 94,8% vs 70% e pT3: 5,2% vs 30%; p=0,005). Correlacionando-se o achado de MCP e estadiamento patológico, observou-se que a maioria das MCP no grupo 1 ocorreu no estadio pT2 (75%), ao passo que 77% das MCP no grupo 2 ocorreu no estadio pT3. Conclusões: O acesso transperitoneal mostrou-se mais eficiente que o acesso extraperitoneal para a realização da prostatectomia radical laparoscópica durante a curva de aprendizado, porém enfatizando que a taxa de complicações graves foi maior quando este acesso foi utilizado. / Introduction: The learning curve in laparoscopic radical prostatectomy (LRP) can vary from 10 to 150 procedures. This procedure can be done using the transperitoneal or the extraperitoneal approach. So far, there is no consensus about the best way to perform LRP, mainly during the initial phases of the LRP programs. Objectives: To analyze and compare the perioperative, oncological and functional results obtained with both approaches while performing LRP during the learning curve. Patients and Methods: Data of the first 40 transperitoneal LRP (Group 1) performed at Getúlio Vargas Hospital of Recife were compared with the first 40 extraperitoneal LRP (Group 2) performed at Clinics Hospital of State University of São Paulo. Results: On transoperative time, statistically significant difference were observed comparing groups 1 and 2 related to the preservation of the neurovascular bundles (77,5% x 90%; p=0,008), overall surgical time (175 min x 267,6 min; p<0,001) and estimated blood loss (177,5 ml x 292,4 ml; p<0,001). Two complications (5%) were observed in group 1 (bleeding and rectal injury), whereas four (10%) were seen in group 2 (bleeding- 5%, rectal and bladder injury). Open conversion occurred in one case (2,5%) in both groups. On postoperative time, statistical difference comparing the groups 1 and 2 were seen in the in-hospital time, indwelling catheter time, narcotic use, biochemical recurrence and mean follow-up time. No statistical difference was observed related to the incidence of early complications (17,5% vs 17,5%; p= 1,000), but three major complications occurred in group 1, leading to one death in this group. On late postoperative time, a statistical difference was observed in the incidence of minor complications (30% vs 15%; p=0,004). Comparing the oncological results between groups 1 and 2, statistical difference was observed in the incidence of positive surgical margins (10,3% vs 32,5%; p=0,016) and pathological stages (pT2: 94,8% vs 70% and pT3: 5,2% vs 30%; p=0,005). The majority of positive margins in group 1 occurred in pT2 (75%), while this observation was more prevalent in pT3 (77%) in group 2. Conclusions: The transperitoneal approach was more efficient than the extraperitoneal approach for performing laparoscopic radical prostatectomy during the learning curve, but major complications were commoner when this approach was adopted.

Pseudomyxoma peritonei (PMP) is a rare disease characterized by mucinous peritoneal metastasis (PM). Different loco-regional treatment strategies, i.e. debulking surgery and cytoreductive surgery (CRS) in combination with hyperthermic intraperitoneal chemotherapy (HIPEC), have changed the prognosis for these patients. CRS is an aggressive surgical procedure with a long learning curve. PMP exists in different types; how many depends on which classification is used. The aims of this thesis were to investigate the time-frame of PMP development from an isolated appendiceal neoplasm, examine the learning process for CRS, evaluate the differences in treatment outcome between debulking surgery and CRS in combination with HIPEC, to evaluate a more detailed PMP classification and to investigate particularly interesting new cysteine-histidine (PINCH) protein as a prognostic factor for PMP. Retrospectively 26 PMP patients were identified as having had an appendectomy with a neoplasm in the appendix but with no evidence of PM at the appendectomy. They were treated for PMP within a median of 13.1 months (3.8-95.3) after the appendectomy. No difference was seen between the types of PMP regarding the time to a clinically significant development of PMP and how much tumour was found at treatment. CRS is a highly invasive treatment and stabilization in the learning curve was seen after 220±10 procedures. Patients treated with CRS+HIPEC had a better 5-year overall survival (OS) than patients treated with debulking surgery, 74% vs. 40%. CRS increased the rate of complete cytoreduction from 25% in patients treated with debulking surgery to 72%. The new four-grade PMP classification showed very good inter-rater agreement between two independent pathologists and a difference in survival rates was observed between the different grades. A positive PINCH staining was recorded in 83% of the tumours and that was associated with poorer survival.

Thesis (Ph. D.)--University of Washington, 2008. / Vita. Includes bibliographical references (leaves 126-133).